1. tf.keras实现线性回归(1)
1.1 Income数据导入可视化
import tensorflow as tf
import pandas as pd
import matplotlib.pylab as plt
print('Tensorflow Version: {}'.format(tf.__version__))
data = pd.read_csv(r'C:\Users\lihuanyu\Desktop\数据集\Income1.csv')
plt.scatter(data.Education, data.Income)
plt.show()
data.head()
Tensorflow Version: 2.0.0
Unnamed: 0 |
Education |
Income |
|
0 |
1 |
10.000000 |
26.658839 |
1 |
2 |
10.401338 |
27.306435 |
2 |
3 |
10.842809 |
22.132410 |
3 |
4 |
11.244147 |
21.169841 |
4 |
5 |
11.645485 |
15.192634 |
1.2 构建模型
data1 = data.to_numpy()
x = data1[:,1]
y = data1[:,2]
model = tf.keras.Sequential()
model.add(tf.keras.layers.Dense(1,input_shape=(1,)))
#查看模型
model.summary()
model.compile(optimizer='adam',loss="mse")Model: "sequential"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
dense (Dense) (None, 1) 2
=================================================================
Total params: 2
Trainable params: 2
Non-trainable params: 0
_________________________________________________________________1.3 训练
history = model.fit(x,y,epochs=10000)
plt.scatter(x,model.predict(x))
plt.scatter(data.Education, data.Income)
plt.show()
2. 多层感知机
2.1 概述
单层神经元的缺陷是神经元要求数据必须是线性可分的，而异或问题无法找到一条直线分割两个类。
为了继续使用神经网络解决这种不具备线性可分性的问题，采取在神经网络的输入端和输出端之间插入更多的神经元。
2.2 数据准备
import tensorflow as tf
print('Tensorflow Version: {}'.format(tf.__version__))
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline
data = pd.read_csv(r'C:\Users\lihuanyu\Desktop\数据集\Advertising.csv')
data1 = data.to_numpy()
data.head()Tensorflow Version: 2.0.0
Unnamed: 0 |
TV |
radio |
newspaper |
sales |
|
0 |
1 |
230.1 |
37.8 |
69.2 |
22.1 |
1 |
2 |
44.5 |
39.3 |
45.1 |
10.4 |
2 |
3 |
17.2 |
45.9 |
69.3 |
9.3 |
3 |
4 |
151.5 |
41.3 |
58.5 |
18.5 |
4 |
5 |
180.8 |
10.8 |
58.4 |
12.9 |
2.3 观察数据分布
plt.scatter(data.TV, data.sales)
plt.show()
plt.scatter(data.radio, data.sales)
plt.show()
plt.scatter(data.newspaper, data.sales)
plt.show()
2.4 建立模型及结果分析
x = data.iloc[:,1:-1].to_numpy()
y = data.iloc[:,-1].to_numpy()
model = tf.keras.Sequential([tf.keras.layers.Dense(10,input_shape=(3,),activation="relu"),
tf.keras.layers.Dense(1)]
)
model.compile(optimizer="adam",loss="mse")
model.summary()
model.fit(x,y,epochs=100)Model: "sequential_5"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
dense_10 (Dense) (None, 10) 40
_________________________________________________________________
dense_11 (Dense) (None, 1) 11
=================================================================
Total params: 51
Trainable params: 51
Non-trainable params: 0
_________________________________________________________________
Train on 200 samples
Epoch 1/100
200/200 [==============================] - 0s 2ms/sample - loss: 759.2462
.......省略
200/200 [==============================] - 0s 122us/sample - loss: 2.1529
Epoch 97/100
200/200 [==============================] - 0s 130us/sample - loss: 2.1491
Epoch 98/100
200/200 [==============================] - 0s 100us/sample - loss: 2.1935
Epoch 99/100
200/200 [==============================] - 0s 170us/sample - loss: 2.1596
Epoch 100/100
200/200 [==============================] - 0s 145us/sample - loss: 2.1212print(model.predict(x[0:5,:]),"\n",y[0:5].reshape(-1,1))[[21.708485 ]
[11.457169 ]
[ 9.0395775]
[18.80669 ]
[12.999142 ]]
[[22.1]
[10.4]
[ 9.3]
[18.5]
[12.9]]
# 3 . softmax多分类任务3.1 任务概述
采用全连接神经网络，实现对Fashion MNIST的分类。Fashion MNIST 数据集包含 70000 张灰度图像，涵盖 10个类别。我们将使用 60000 张图像训练网络，并使用 10000 张图像评估经过学习的网络分类图像的准确率。可以从 TensorFlow 直接访问 Fashion MNIST，只需导入和加载数据即可。
3.2 数据准备
import tensorflow as tf
print('Tensorflow Version: {}'.format(tf.__version__))
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline
#图像，标签，测试图像与标签
(train_image, train_lable), (test_image, test_label) = tf.keras.datasets.fashion_mnist.load_data()
print("train_image shappe:{}\ntest_image shape:{}".format(train_image.shape,test_image.shape))
plt.imshow(train_image[0])
plt.axis("off")
plt.show()
#归一化
train_image = train_image/255
test_image = test_image/255Tensorflow Version: 2.0.0
train_image shappe:(60000, 28, 28)
test_image shape:(10000, 28, 28)
3.3 建立模型并训练
model = tf.keras.Sequential()
model.add(tf.keras.layers.Flatten(input_shape=(28,28)))
model.add(tf.keras.layers.Dense(128,activation="relu"))
model.add(tf.keras.layers.Dense(10,activation="softmax"))
model.summary()
model.compile(optimizer = "adam",loss='sparse_categorical_crossentropy',metrics=['acc'])
model.fit(train_image,train_lable,epochs=5)Model: "sequential_2"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
flatten_2 (Flatten) (None, 784) 0
_________________________________________________________________
dense_4 (Dense) (None, 128) 100480
_________________________________________________________________
dense_5 (Dense) (None, 10) 1290
=================================================================
Total params: 101,770
Trainable params: 101,770
Non-trainable params: 0
_________________________________________________________________
Train on 60000 samples
Epoch 1/5
60000/60000 [==============================] - 8s 138us/sample - loss: 3.1367 - acc: 0.6865
Epoch 2/5
60000/60000 [==============================] - 8s 130us/sample - loss: 0.6911 - acc: 0.7243
Epoch 3/5
60000/60000 [==============================] - 7s 114us/sample - loss: 0.6168 - acc: 0.7527
Epoch 4/5
60000/60000 [==============================] - 8s 128us/sample - loss: 0.5736 - acc: 0.7814
Epoch 5/5
60000/60000 [==============================] - 7s 117us/sample - loss: 0.5319 - acc: 0.8114#评价
model.evaluate(test_image, test_label)10000/1 - 1s 69us/sample - loss: 0.3796 - acc: 0.8244
[0.5249296184539795, 0.8244]3.5 采用one-hot编码方式训练
train_label_onehot = tf.keras.utils.to_categorical(train_lable)
test_label_onehot = tf.keras.utils.to_categorical(test_label)
model = tf.keras.Sequential()
model.add(tf.keras.layers.Flatten(input_shape=(28,28))) # 28*28
model.add(tf.keras.layers.Dense(128, activation='relu'))
model.add(tf.keras.layers.Dense(10, activation='softmax'))
model.summary()
model.compile(optimizer=tf.keras.optimizers.Adam(learning_rate=0.001),
loss='categorical_crossentropy',
metrics=['acc']
)
model.fit(train_image, train_label_onehot, epochs=5)Model: "sequential_3"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
flatten_3 (Flatten) (None, 784) 0
_________________________________________________________________
dense_6 (Dense) (None, 128) 100480
_________________________________________________________________
dense_7 (Dense) (None, 10) 1290
=================================================================
Total params: 101,770
Trainable params: 101,770
Non-trainable params: 0
_________________________________________________________________
Train on 60000 samples
Epoch 1/5
60000/60000 [==============================] - 9s 148us/sample - loss: 3.9736 - acc: 0.7085
Epoch 2/5
60000/60000 [==============================] - 7s 122us/sample - loss: 0.6800 - acc: 0.7587
Epoch 3/5
60000/60000 [==============================] - 8s 135us/sample - loss: 0.6008 - acc: 0.7850
Epoch 4/5
60000/60000 [==============================] - 7s 119us/sample - loss: 0.5508 - acc: 0.8023
Epoch 5/5
60000/60000 [==============================] - 8s 125us/sample - loss: 0.5175 - acc: 0.8154
<tensorflow.python.keras.callbacks.History at 0x2fe4bf91388>#评价
model.evaluate(test_image, test_label_onehot)1s 72us/sample - loss: 0.3707 - acc: 0.8031
[0.5642493257284165, 0.8031]
3.6 多添加几层隐藏层
model = tf.keras.Sequential()
model.add(tf.keras.layers.Flatten(input_shape=(28,28))) # 28*28
model.add(tf.keras.layers.Dense(128, activation='relu'))
model.add(tf.keras.layers.Dense(128, activation='relu'))
model.add(tf.keras.layers.Dense(128, activation='relu'))
model.add(tf.keras.layers.Dense(10, activation='softmax'))
model.summary()Model: "sequential_2"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
flatten_2 (Flatten) (None, 784) 0
_________________________________________________________________
dense_4 (Dense) (None, 128) 100480
_________________________________________________________________
dense_5 (Dense) (None, 128) 16512
_________________________________________________________________
dense_6 (Dense) (None, 128) 16512
_________________________________________________________________
dense_7 (Dense) (None, 10) 1290
=================================================================
Total params: 134,794
Trainable params: 134,794
Non-trainable params: 0
_________________________________________________________________model.compile(optimizer="adam",
loss='categorical_crossentropy',
metrics=["acc"])
#加入验证集
history = model.fit(train_image,train_label_onehot,epochs=20,validation_data=(test_image,test_label_onehot))
plt.plot(history.epoch,history.history.get("acc"),label="train_acc")
plt.plot(history.epoch,history.history.get("val_acc"),label="val_acc")
plt.legend()
plt.show()
从图中，我们可训练集的准确率，在不断上升，而验证集的数据没有上升，甚至还有些下降，这是典型的过拟合状态，常见的，我们可以采用丢弃法或者减少层数来抑制过拟合状态。
3.7 添加dropout
model = tf.keras.Sequential()
model.add(tf.keras.layers.Flatten(input_shape=(28,28))) # 28*28
model.add(tf.keras.layers.Dense(128, activation='relu'))
model.add(tf.keras.layers.Dropout(0.5))
model.add(tf.keras.layers.Dense(128, activation='relu'))
model.add(tf.keras.layers.Dropout(0.5))
model.add(tf.keras.layers.Dense(128, activation='relu'))
model.add(tf.keras.layers.Dropout(0.5))
model.add(tf.keras.layers.Dense(10, activation='softmax'))
model.compile(optimizer="adam",
loss='categorical_crossentropy',
metrics=["acc"])
#加入验证集
history = model.fit(train_image,train_label_onehot,epochs=20,validation_data=(test_image,test_label_onehot))
plt.plot(history.epoch,history.history.get("acc"),label="train_acc")
plt.plot(history.epoch,history.history.get("val_acc"),label="val_acc")
plt.legend()
plt.show()
有上图我们可以发现，验证集的准确度比训练集的准确度高，所以dropout具有抑制过拟合的效果。